Grief trajectories and long-term health effects in bereaved relatives: a prospective, population-based cohort study with ten-year follow-up.

Background: Bereavement may affect the health of relatives, causing increased use of health care services and increased mortality shortly after the patient's death. However, the long-term consequences for those with a high level of grief symptoms remain largely unexplored. We aimed to investigate associations between grief symptom trajectories and four long-term health outcomes among relatives bereaved by natural death: contacts to general practice and mental health services, use of psychotropic prescription medication, and mortality, over a period of 3-10 years post-bereavement. Method: We assessed grief symptoms using the Prolonged Grief-13 scale in a cohort of 1,735 bereaved relatives at three different time points (prior to bereavement, 6 months after bereavement, and 3 years after bereavement) and identified five main grief trajectories. The trajectory with persistent low levels of grief symptoms in relatives [n=670 (45%)] was called the low grief trajectory (LGT) and was used as reference. The high grief trajectory (HGT) consisted of 107 (6%) relatives with persistent high grief symptom levels. We investigated associations between grief trajectories and (1) contacts to general practitioner (GP) including out-of-hours using negative binomial regression analysis, (2) contacts to mental health services (GP talk therapy, private-practice psychologist or psychiatrist), (3) use of psychotropic medications (antidepressants, anxiolytics and sedatives) using logistic regression analysis, and (4) mortality using Cox regression analysis. The follow-up period started at 3 years after bereavement and long-term outcome were further followed until 10 years after the patient's death. Results: Relatives in the HGT had a significantly higher yearly incidence of GP contacts until seven years after bereavement compared to the LGT. The HGT was associated with higher use of mental health services [OR = 2.86 (95% CI 1.58; 5.19)], antidepressants [OR = 5.63 (95% CI 3.52; 9.01)], sedatives and anxiolytics [OR = 2.60 (95% CI 1.63; 4.14)], and excess mortality [HR = 1.88 (95% CI 1.1; 3.2)] compared to the LGT. Conclusion: This study shows that patients with high and sustained grief symptoms have an increased healthcare use up to 10 years after loss. Future research should assess whether current health care services sufficiently meet the prolonged needs of these relatives.